Fog Creek Software
Discussion Board

I don't save email for 5 years...

I don't save email for 5 years but then again I don't run a business.  I'm wondering how long people save their emails?  I have a hard time saving anything without immediate value to me.  If I download something interesting off the internet I burn it to a CD.  I guess it would make sense to store all of this stuff somehow in an easily searchable format.  When I recieve an email I read it, respond and delete when all correspondence is done for that message.  I hate keeping emails (or anything else) hanging around.  If I were to run a business then sure I'd keep the emails but how long is too long (if there is such a thing)?

Monday, July 19, 2004

>  easily searchable format

It isn't easy. That's why Joel wrote an article recently about a tool to fixing Outlook's search.

Li-fan Chen
Monday, July 19, 2004

It's hard to search through zip files on file systems without the proper tools, it's even harder if the files are in a mail box--another worthwhile consideration.

Li-fan Chen
Monday, July 19, 2004

An interesting article about email and searching:
(Watching an ad is required to read the entire article).

I don't think there is any reason every single word in several years of email cannot be quietly indexed in the background and searched as you type the keywords.  I've seen other mail clients that do this wonderfully on tens of thousands of emails.  I hope the predictions of this article are true and mainstream email clients go this way.

Monday, July 19, 2004

I guess I was referring to the fact that Joel kept his email for five years.  This is probably for business purposes though.  How long do you guys normally keep your emails?

Monday, July 19, 2004

I have every email I've ever sent since 1996.  I have pretty much every email I've ever received since like 1997-98.  I use Eudora, which scales excellently and the search works extremely well.

My mail is sorted in folders (one for each employer, then one for each client inside that).  And I never really thing about it.

Why would I delete my mail?

Almost Anonymous
Monday, July 19, 2004

Business email : 3 years and counting. (I switched email programs 3 years ago).

Personal: 6 months or so.

Mr. Analogy
Monday, July 19, 2004

I've kept all my email since 1995 (freshman year in college). Basically, the only things I clear out are spam and insignificant one-off emails like "meeting tomorrow at 5pm".

Monday, July 19, 2004

Never consciously "keep" email, just don't bother to delete.

Tuesday, July 20, 2004

Does anyone think it's feasible to use GMail for business email? The searching in GMail I'm sure is way on par with what Joel is talking about, apart from the whole thing being online of course. I haven't been able to look at GMail myself though because I don't have an invite? Anyone got a spare one they might like to donate? I'd love to check it out! :D

Nathan Ridley
Tuesday, July 20, 2004

I save every e-mail I ever get.  The only reason I ever lose mail is because of disk failure (usually every few years ;)).

If I was more together, I would back it up to CD.

Jimmy Jo-jo
Tuesday, July 20, 2004

You know, maybe I just don't "get" this, but what is all this fuss about ?

I run two ISV businesses and have been using email for about 12 years. Conveniently, I have set up rules in Outlook so that email from various people goes into the correct folder.

If I need to find an email I just look in the appropriate folder and there it is. I don't need a stupid search engine.

I admit that if I wanted to find all emails with the word "cat" in them, more folder-based approach would not be ideal, but I don't want to search like that.

Steve Jones (UK)
Tuesday, July 20, 2004

"If I need to find an email I just look in the appropriate folder and there it is. I don't need a stupid search engine."

Mildly off-topic, but here's where I really like Gmail's concept of "labels" attached to messages: a single message can have arbitrarily many labels, but it usually only exists in one folder.  It lets me deal with messages that fit more than one subject, and/or orthogonal classification schemes (one set of labels for sender, another for subject).  Of course, if you're just labelling by sender, you're probably wasting your time, because search by sender is easy, but that's the idea.

If you restate the problem of classifying emails into folders as "what future searches will I want to include this message?", that job gets pretty difficult.  I don't know what I'll be doing a year from now, so it's hard for me to determine what of today's email will be relevant to it.  In a business scenario, if you're dealing with the process of discovery (turn over all emails related to "Enron", for example), this problem gets much bigger.

Tuesday, July 20, 2004

I keep all my mail: via procmail and a Perl script, all non-spam, non-mailing-list mail gets dropped off into three folders: INBOX, archive-all, and archive-yyyymm. That way, i can delete messages out of my INBOX when i'm done with them and know they're archived, rather than having to save them from INBOX to the archive.

I use a program called mutt to read my mail. It has extensive and very fast search capabilities, plus i save my mail in maildir format: each folder is a directory, each message is its own file. So i could always just use grep to search my messages too. And splitting them the archive by month makes searches even more managable.

If anyone here runs a Unix variant and wants to see the code / config files, let me know. (Hunting down my email address is left as an excercise for the reader)

Mike Schiraldi
Tuesday, July 20, 2004

I keep all (ie. what I haven't deleted) my email in between changing mail clients.

I change mail clients about every 3-5 years.

If you asked me about a mail in 1993 I could probably go find it, but I would hate you unto the seventh generation and if I ever found your grave I would dance upon it and sing comic songs.

Simon Lucy
Tuesday, July 20, 2004

On-topic: I have all my personal email from when I first got online at home (circa 1998 or so).

Slightly off-topic: That personal email, and some mail archives I've saved from jobs I've had along the way, is all stored in Outlook/Outlook Express archives.  (the personal stuff under MacOS 8.6.1, the rest NT/2k). 

I'd like to consolidate all this into one format organized onto one disk (probably on my little Linux server & available to my Mac, Wintel, and Linux boxen).  During a discussion of email clients & format conversion a while back (can't find the thread) someone mentioned the "one true email format" (defined by some RFC).  It seems like converting everything to this format would give me maximum flexibility re: choosing a mail client (and which platform I want to run it under) in the future.  If anyone can share any experience with this it'd be much appreciated - thanks.

- former car owner in Queens
Tuesday, July 20, 2004

(realize I should have phrased my question a bit more clearly)

Will the effort required to convert all my mail archives to one standard format pay off by giving me the flexibility to freely change email clients?  Opinions/experiences appreciated...

- former car owner in Queens
Tuesday, July 20, 2004

theres a feature in outlook that sounds somewhat similar to labels in gmail... searchfolders or something like that. something like virtual folders...

Tuesday, July 20, 2004

==>I'm wondering how long people save their emails

I have every email I ever sent or received (less the spam, of course) since my first "real" job in 1994. Prior to that, I have about half of the email sent/received during college (1990-1994). The other (missing) half was the result of a hard drive failure combined with flaky backups. Who wants the old correspondance from college anyway?

To date, it's just over 9GB of email (plus attachments). FYI -- I never search for anything older than 2 years old, but I've still got it in the archives JustInCase.

Sgt. Sausage
Tuesday, July 20, 2004

I use a program called Nelson Email Organizer which seems to solve the Outlook searching issue. It just sits on top of Outlook (and doesn't touch your main outlook data store so you can get rid of it at any time). It indexes all emails and searches happen almost instantly, and it also sorts email into a whole variety of folders - so for example one email might appear in the 'today', 'client x', 'with attachment' and 'high priority' folders all at once. I have found it very useful.

Well worth a (and no, I don't have any kind of relationship with the company. Just a satisfied customer!)

I, like a previous poster, just never actually delete any emails except spam. I just don't see the point - it's not like disk space is a problem, and I have often needed to refer back to old emails for one reason or another.

James U-S
Tuesday, July 20, 2004

IANAL but there are legal issues with respect to e-mail retention.  Did no one learn from the Microsoft-DOJ lawsuit?

Tuesday, July 20, 2004

I have my email going back to 1998 on my hard disk, and a few years farther back than that on CD and Zip.

It's all in text files - roughly one per month.  It's easily searchable with a variety of tools - I use Ultraedit because its "find in files" works for me.  I used to go through and remove any attached files, but haven't bothered lately.

Tuesday, July 20, 2004

"I admit that if I wanted to find all emails with the word "cat" in them, more folder-based approach would not be ideal, but I don't want to search like that."

What about these:
  "all the emails relating to this project"
  "all the emails relating to this project from that person"
  "all the emails relating to this project's deadline"
  "all the emails containing contracts for this project"

Are you saying that wouldn't be useful? Or are you saying that it's not currently realistic so it's just not the way you work right now?

I'm quite happy with the idea that some people would actually prefer not to have anything other than a simple set of folders to store email in, but I suspect most people would find more powerful systems to actually be quite useful. (This is, of course, assuming that it's not excessively difficult to use - but that's a whole nother story.)

Tuesday, July 20, 2004

"IANAL but there are legal issues with respect to e-mail retention.  Did no one learn from the Microsoft-DOJ lawsuit?"

We learnt that if you're breaking the law you should cover your ass more carefully.

We also learnt that if you're sufficiently wealthy then "but that would hurt" becomes a valid argument against punishment for doing things that, er, um, the law says you're not allowed to do.

So, overall, the plan is to become so wealthy that the law becomes optional guidelines, at which point the email retention doesn't hurt. Got it.  :)

Tuesday, July 20, 2004

I hire people for our college. I have correspondence from over 3,000 applicants, and keep all email in a folder with at present 18,000 mails in it.

I often need to check all old correspondence, particularly when somebody writes back. I open Contacts and then the Activities tab, so I get all messages to and from the guy's various email addresses. This takes ages and because Outlook is single-threaded stops me doing anything else.

As Joel says this is idiotic. The one thing you want to search all the time, and thus have indexed, is your email folder. And this is the one thing that MS doesn't offer to index.

Stephen Jones
Tuesday, July 20, 2004

*  Recent Topics

*  Fog Creek Home